Adaptive Grid Computing

1
Adaptive Grid Computing

• Detecting and Adjusting for Dynamic Grid
  Conditions

2
Experiments with a Grid-Enabled Computational
Framework

• Dave Angulo, Ian Foster
• Chuang Liu, Matei Ripeanu, Michael Russell,
  Lingyun Yang
• Distributed Systems Laboratory
• University of Chicago Argonne National
  Laboratory
• Gabrielle Allen, Thomas Dramlitsch,
• Gerd Lanferman, Ed Seidel, Thomas Radke
• Max-Planck-Institut für Gravitationsphysik

3
Acknowledgements

• This work was supported in part by the NSF-funded
  Grid Application Development Software project
  under Grant No. 9975020.
• We are grateful to our GrADS project colleagues
  for discussions on the topics discussed here.

4
Research Goals

• Investigate methods and structures for efficient
  Grid execution via in-depth study of a demanding
  application, including
• Constructs for adapting to heterogeneity
• Constructs for dynamic resource acquisition
• Create testbed for GrADSoft components, as they
  emerge
• Investigate utility of computational frameworks
  as facilitator of Grid computing

5
Adaptation to Dynamic Grid Environments

• Migrate to faster/ cheaper system
• When better system discovered
• When requirements change
• When characteristics change (e.g., competition)

6
Adaptation to Dynamic Grid Environments

• Migrate to faster/ cheaper system
• When better system discovered
• When requirements change
• When characteristics change (e.g., competition)

7
Cactus Worm Architecture
External Processes / Services
CactusWormServer
Globus ToolkitGRAM
Cactus flesh
NWS
Resource selector
GlobusToolkitGIIS
8
Migration Demonstration
GIIS Running on jupitor.isi.edu
Resource Selector Service Running on
hamachi.cs.uchicago.edu
Cluster at UIUC
Cluster at UCSD
Cactus Worm Server Running on amajor.cs.uiuc.edu
9
Migration Demonstration
GIIS
Resource Selector Service
Cluster at UIUC
Cluster at UCSD
Cactus WormServer
Computational application begins run
NetLogger records performance and displays visual
graph
10
Migration Demonstration
GIIS
Resource Selector Service
Cluster at UIUC
Cluster at UCSD
Cactus WormServer
Competing application on one node steals CPU
cycles
NetLogger
11
Migration Demonstration
GIIS
Resource Selector Service
Cluster at UIUC
Cluster at UCSD
Cactus WormServer
Resource Selector Service contacted
NetLogger
12
Migration Demonstration
GIIS
Resource Selector Service
Cluster at UIUC
Cluster at UCSD
Cactus WormServer
resource list sent to Worm Server
NetLogger
13
Migration Demonstration
GIIS
Resource Selector Service
Cluster at UIUC
Cluster at UCSD
Cactus WormServer
Worm Server shuts down app
NetLogger
14
Migration Demonstration
GIIS
Resource Selector Service
Cluster at UIUC
Cluster at UCSD
Cactus WormServer
Worm Server starts app on new resouces
NetLogger
15
Performance Results
16

• Gabrielle Allen Dave Angulo Ian Foster Gerd
  Lanfermann Chuang Liu Thomas Radke Ed Seidel
  John Shalf. The Cactus Worm Experiments with
  Dynamic Resource Discovery and Allocation in a
  Grid Environment. In International Journal of
  High-Performance Computing Applications Volume
  15, Number 4, 2001
• http//people.cs.uchicago.edu/dangulo/papers/
• This presentation http//www.cs.uchicago.edu/dan
  gulo/scdemo.ppt